sign in sign up
<<
Home , Manicina, Manicina areolata, Mussidae, Pectiniidae, Scleractinia

Schwartz et al. BMC Evolutionary Biology 2012, 12:123 http://www.biomedcentral.com/1471-2148/12/123

RESEARCH ARTICLE Open Access Molecules and reveal punctuated diversification in “faviid” Sonja A Schwartz1*, Ann F Budd2 and David B Carlon3

Abstract Background: Even with well-known sampling biases, the record is key to understanding macro-evolutionary patterns. During the Miocene to Pleistocene in the Caribbean Sea, the fossil record of scleractinian corals shows a remarkable period of rapid diversification followed by massive extinction. Here we combine a time-calibrated molecular phylogeny based on three nuclear introns with an updated fossil stratigraphy to examine patterns of radiation and extinction in Caribbean corals within the traditional Faviidae. Results: Concatenated phylogenetic analysis showed most of Caribbean faviids were monophyletic, with the exception of two species. The time-calibrated tree revealed the stem group originated around the closure of the Tethys Sea (17.0 Ma), while the Manicina diversified during the Late Miocene (8.20 Ma), when increased sedimentation and productivity may have favored free-living, heterotrophic species. and shallow water specialists, represented by and , originate at the beginning of the Pliocene (5 – 6 Ma) as the Isthmus of Panama shoaled and regional productivity declined. Conclusions: Later origination of the stem group than predicted from the fossil record corroborates the hypothesis of morphological convergence in Diploria and Favia genera. Our data support the rapid evolution of morphological and life-history traits among faviid corals that can be linked to Mio-Pliocene environmental changes. Keywords: , Speciation, Adaptive radiation, Miocene, Pliocene, reef

Background historical information from both molecular and fossil Explaining rapid diversification and speciation remains a data. By examining systems that show recent speciation central challenge to evolutionary biology [1,2]. Much within monophyletic groups, ecological differentiation, work has focused on either understanding the ecology and a strong fossil record, we can begin to link past to and phylogenetic history of species-rich systems that present processes in the understanding of the evolution have recently diversified along ecological axes (e.g. adap- of diversity. tive radiations) [3], or looking for patterns of species The marine Caribbean fauna provides rare examples of change in the fossil record [4-8]. Taking the molecular diversification of monophyletic lineages within the con- phylogenetic/ecological approach alone, however, text of well-understood changes in biogeography, ocean- excludes information about extinct lineages that may ography, and climate. The isolation of Caribbean substantially bias our ability to identify cases of rapid di- populations from their Indo-Pacific counterparts started versification [9]. Conversely, relying on the fossil record ~15-17 Ma when the closure of the Tethys Sea cut off alone limits our ability to detect evolutionary relation- connections between the Mediterranean and Indo- ships between fossil taxa and some shifts in ecological Pacific [10]. Isolation was complete ~3.45 - 4.25 Ma function that may not be apparent from fossil character when the rise of the Isthmus of Panama severed all states. Ultimately, a more complete understanding of Caribbean connections to the Indo-Pacific [11]. The the processes that drive rapid diversification will require period leading up to closure of the isthmus during the late Miocene to late Pliocene was characterized by chan- * Correspondence: [email protected] ging global oceanographic circulation patterns, leading 1Department of Environmental Science, Policy & Management, University of to drastic environmental, ecological, and taxonomic California, Berkeley, CA 94720, USA shifts within the Caribbean basin. Not only did the Full list of author information is available at the end of the article

© 2012 Schwartz et al.; licensee BioMed Central Ltd. This is an Open Access article distributed under the terms of the Creative Commons Attribution License (http://creativecommons.org/licenses/by/2.0), which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited. Schwartz et al. BMC Evolutionary Biology 2012, 12:123 Page 2 of 10 http://www.biomedcentral.com/1471-2148/12/123

cessation of gene flow between the Pacific and Atlantic demands an independent assessment of trends apparent Oceans lead to widespread vicariant speciation across in the fossil record. the newly formed isthmus [12-14], but on the Caribbean To explore the tempo and mode of this evolutionary di- side, the accompanying geological and oceanographic versification, we unite a new multi-locus phylogeny of the changes caused an overall decrease in depth, primary Caribbean Faviidae with new stratigraphic compilations productivity and turbidity and an increase in salinity, from the fossil record. Our well-sampled phylogeny temperature, and local environmental heterogeneity allows Bayesian approaches to place these relationships [11,15]. Fossil records of many marine taxa during this into a temporal context by dating divergence times based period show elevated levels of taxonomic turnover on molecular data and fossil calibrations. We compare [11,16-21], suggesting that climatic and geological vari- our time-calibrated phylogeny to temporal patterns of ables drove elevated rates of cladogenesis and extinction. origination and extinction revealed by the Neogene fossil This taxonomic turnover is particularly striking in cor- record, and find remarkable congruence between data als of the family Faviidae, where an examination of the sets. The timing of events revealed by this analysis stratigraphic ranges shows that all extant species origi- strongly implicates paleoenvironmental changes as dri- nated nearly simultaneously during the Mio-Pliocene vers of diversification in scleractinian corals. [22]. Moreover, for faviids, this recent radiation has resulted in impressive diversification of ecological and Results life-history traits [23,24]. Modern species of Manicina Phylogenetic analysis of Caribbean “Faviidae” are representative of a free-living lifestyle adapted to We sequenced three single copy nuclear loci for six sediment-rich seagrass habitats that expanded during the ingroup and one outgroup Caribbean faviid species. A Miocene then contracted during the Plio-Pleistocene total of 48 unique alleles were identified for CaM (align- [15]. In contrast, species of the genus Diploria ment length = 507 bp), 38 alleles were identified for tend to be reef-builders, dominating shallow water reef MaSC-1 (alignment length = 490 bp), and 55 alleles were platforms in Pleistocene and modern times [25-28]. identified for Pax-C (alignment length = 418 bp) (Add- These two “sediment” and “reef” clades appear to share a itional file 1). Maximum likelihood and Bayesian analysis common ancestor and ecological diversification seems to of gene trees showed little support for structure above have occurred over a short period of geological time, sug- the species level with no conflict between trees at highly gesting it is tied to the contemporaneous increase in en- supported nodes (Additional file 2). The taxa Manicina vironmental heterogeneity [29]. Yet this punctuated areolata and M. mayori shared some alleles at all loci, diversification event is inferred from a fossil record, and unique alleles isolated from Diploria clivosa and D. which may be incomplete or contain uncertainties in dat- strigosa did not always form monophyletic groups. A ing and taxonomic relationships that may influence our total of 94 individuals with unique genotypes were suc- interpretation of past patterns. cessfully sequenced at all three loci and used for a conca- Molecular data combined with well sampled fossil tenated phylogenetic analysis. See Additional file 3 for records provide opportunities to test existing evolution- genotype data of all individuals in study. ary hypotheses and extend our understanding of both Bayesian and maximum likelihood trees had identical the tempo and mode of evolutionary diversification. In topologies at all major nodes with support values the Scleractinia, deep divergences between coral orders, (Bayesian/ML bootstrap) indicated in Figure 1. The suborders and families are increasingly well understood ingroup node was well supported (100/100) as well as [30-33]. Yet a recent series of phylogenies exploring rela- species nodes for C. natans (100/100), D. clivosa (100/100), tionships at the familial level and below have demon- D. labyrinthiformis (100/98), D. strigosa (100/98) and strated pervasive polyphyly and paraphyly at the generic F. fragum (100/100). Manicina mayori and Manicina level [34-39]. In addition, these studies have shown that areolata failed to form monophyletic groups, though between ocean basins, species group geographically ra- support was high at the genus node for Manicina (100/ ther than taxonomically [35,38,39]. In particular, Atlantic 94). The genus Diploria failed to form a monophyletic lineages of Faviidae and appear to be more group. Diploria clivosa formed a clade with Manicina closely related to other Atlantic lineages than to conge- spp. and D. strigosa formed a clade with . ners or even confamilials in other ocean basins. This Support for these nodes, however, was low (72/68 and geographic split supports the evidence from the fossil 76/60 respectively). record of a radiation of the Caribbean coral fauna before complete isolation from the Pacific. However, the failure Timing of divergence to resolve species relationships within the traditional BEAST analysis of the data produced a tree topologically coral family Faviidae, and a long history of taxonomic consistent with those of the MrBayes and RaxML ana- difficulties in identifying and classifying corals [32,36,40] lyses. Visual inspection of plots in Tracer v1.5 [41] Schwartz et al. BMC Evolutionary Biology 2012, 12:123 Page 3 of 10 http://www.biomedcentral.com/1471-2148/12/123

100/100 C. natans 1019 (P) 100/100 C. natans 1017 (P) C. natans 521 (P) D. clivosa 1043 (P) D. clivosa 1044 (P) D. clivosa 1046 (P) D. clivosa 1050 (P) D. clivosa 1338 (F) 100/100 D. clivosa 1355 (F) D. clivosa 1356 (F) D. clivosa 1357 (F) D. clivosa 833 (P) D. clivosa 834 (P) D. clivosa 835 (P) D. clivosa 998 (P) M. areolata 1320 (F) M. areolata 1321 (F) M. areolata 1322 (F) M. areolata 1323 (F) M. areolata 1324 (F) * M. areolata 1325 (F) Diploria clivosa M. areolata 1326 (F) M. areolata 982 (P) M. areolata 988 (P) M. areolata 992 (P) M. areolata 986 (P) M. areolata 990 (P) M. areolata 993 (P) -/81 M. areolata 989 (P) M. mayori 1035 (P) 100/94 M. mayori 1036 (P) M. mayori 1037 (P) M. mayori 1034 (P) M. areolata 987 (P) M. areolata 1327 (F) M. areolata 1328 (F) M. areolata 1335 (F) M. areolata 991 (P) M. areolata 1330 (F) M. areolata 1333 (F) M. areolata 1334 (F) M. areolata 1337 (F) 96/- M. areolata 983 (P) M. areolata 985 (P) M. areolata 984 (P) M. mayori 1020 (P) D. labyrinthiformis 1025 (P) D. labyrinthiformis 1026 (P) D. labyrinthiformis 1038 (P) D. labyrinthiformis 1039 (P) D. labyrinthiformis 1040 (P) D. labyrinthiformis 1041 (P) 100/100 D. labyrinthiformis 516 (P) D. labyrinthiformis 853 (P) D. labyrinthiformis 854 (P) D. labyrinthiformis 855 (P) D. labyrinthiformis 856 (P) 100/98 D. labyrinthiformis 857 (P) D. labyrinthiformis 1339 (F) D. labyrinthiformis 1341 (F) Diploria labyrinthiformis D. labyrinthiformis 1351 (F) D. labyrinthiformis 1365 (F) D. labyrinthiformis 1340 (F) D. strigosa 1027 (P) D. strigosa 1030 (P) D. strigosa 1032 (P) D. strigosa 1353 (F) D. strigosa 519 (P) D. strigosa 520 (P) D. strigosa 849 (P) D. strigosa 852 (P) 100/98 D. strigosa 850 (P) D. strigosa 1342 (F) D. strigosa 1363 (F) D. strigosa 1364 (F) D. strigosa 851 (P) Diploria strigosa F. fragum 1103 (P) F. fragum 1305 (F) 100/- F. fragum 1308 (S) F. fragum 1375 (S) F. fragum 1369 (S) Colypophyllia natans F. fragum 1374 (S) ** F. fragum 1390 (S) F. fragum 1378 (S) F. fragum 1382 (S) F. fragum 603 (P) F. fragum 641 (P) F. fragum 755 (P) F. fragum 756 (P) F. fragum 1393 (S) 100/100 F. fragum 514 (P) F. fragum 1383 (S) F. fragum 1386 (S) F. fragum 1387(S)

0.5 substitutions/site Favia fragum Figure 1 and photographs of species of the Caribbean Faviidae. Tree based on a partitioned analysis of individual genotypes at the CaM, MaSC-1, and Pax-C loci. Terminal taxa are individuals of each species. Letters after sample names indicate coarse geographic sampling information (F = , P = Panama, S = St. Croix). Further sampling and genotype information can be found in Additional file 3. Trees shown were created using Bayesian methods in MrBayes v3.1. Maximum likelihood trees created in RaxML yielded a similar topology. Posterior probabilities (>95%) and bootstrap support (>75%) (Bayesian/ML) are indicated for each node. Dashes indicate nodes unsupported in an analysis. Several deeper nodes in the tree indicated by asterisks were poorly supported in this analysis (* = 72/68, ** = 76/60). Photographs of each species show morphological diversity within this clade. All Diploria and species are reef-building, while Favia and Manicina species are also free-living. (Photo credit: Dr. Charles and Anne Sheppard, http://coralpedia.bio.warwick.ac.uk/). showed rapid convergence of the analysis and narrowing Table 1. The posterior mean of the time of the most re- of priors with all parameters having an effective sample cent common ancestor (TMRCA) of the Manicina size (ESS) of >1900. The mean rate of substitution was group, which was calibrated from species fossil data, 6.77× 10-4 per site (95% Highest Posterior Density shifted several MY from the prior distribution, indicating (HPD) interval: 4.49 × 10-4 - 9.06× 10-4) with a coeffi- that the sequence data is influencing divergence dates. cient of variation of 1.17 (95% HPD interval: 0.66- 1.72) For D. clivosa, D. strigosa, and F. fragum, mean esti- indicating significant heterogeneity in substitution rates mated ages fell close to the earliest possible dates of their across the tree. appearance in the fossil record. For the D. labyr- Mean ages of species, ingroup, and root nodes with inthiformis and Manicina nodes, fossil dates were closer 95% HPD intervals are shown in Figure 2A and listed in to the youngest part of the 95% HPD interval. Mean Schwartz et al. BMC Evolutionary Biology 2012, 12:123 Page 4 of 10 http://www.biomedcentral.com/1471-2148/12/123

A. IUGS Stage (Ma)

Manicina Diploria Favia Diploria Diploria C. clivosa fragum strigosa labyrinthiformis natans 0 Gelasian shu fPnm lsr Tethys Sea Closure Isthmus of Panama Closure

2.5 Piac. Zanc.

5.0 Mess.

7.5 Tortonian

10.0 ? ? Serr. 12.5 ? Langhian

15.0 Burdiglian 17.5

Mean Node Age 20.0

95% HPD Interval Aquitanian

Stratigraphic Range of 1st Appearance 22.5 Geological Event

25.0

B.

Present

10 Ma

20 Ma

30 Ma

40 Ma

50 Ma Thysanus corbicula Thysanus exentricus Thysanus navicula Thysanus sp. A Hadrophyllia saundersi Manicina sp. E Manicina aff. mayor Manicina geisteri grandis Manicina jungi Manicina puntagordensis Manicina areolata Manicina pliocenia Manicina Manicina mayor Diploria antiguensisi Diploria dumblei Diploria portoricensis Diploria zambensis Diploria bowersi Diploria clivosa Diploria strigosa Diploria labyrnthiformis Diploria sarasotana Favia favioides* Favia gregoryi Favia weisbordi Favia macdonaldi Favia dominicensis Favia aff. dominicensis Favia maodentrensis Favia vokesae Favia fragum Favia gravida Colpophyllia duncani Colpophyllia elegans Colpophyllia willoughbiensi Colpophyllia mexicanum Colpophyllia sp. A Colpophyllia breviserialis Colpophyllia natans Colpophyllia amaranthus i i i s

Figure 2 (See legend on next page.) Schwartz et al. BMC Evolutionary Biology 2012, 12:123 Page 5 of 10 http://www.biomedcentral.com/1471-2148/12/123

(See figure on previous page.) Figure 2 Caribbean Faviidae chronogram and stratigraphic data. A) Divergence dates of terminal (species) and internal nodes of a phylogeny of the Caribbean Faviidae. Original chronogram and tree generated in BEAST. Grey boxes indicate species or genera as labeled. Black circles and blue bars correspond to mean node age (Ma) and 95% HPD intervals produced by BEAST analysis. Red bars indicate the stratigraphic age range of the first appearance of that in the fossil record. Green bars next to the time axis are used to indicate major geological events in the isolation of the Caribbean Sea including the closure of the Central American Isthmus at 4.25-3.5 Ma and the closure of the Tethys Sea at 17–15 Ma. Nodes marked with a '?' are poorly supported in this analysis. Detailed information about dates and node calibration can be found in Tables 1 and 2. B) Phylogeny on stratigraphy of living and extinct species. Stratigraphic range bars are color-coded by genera, listed on the x axis. Green + blue shading are 95% highest posterior density (HPD) intervals for the ingroup node, and green + yellow shading are 95% HPD intervals for the root node as seen on the chronogram. Orange shading indicates the range of origination dates in the fossil record for all living taxa. Species within genera are ranked by earliest origination date, left to right. The genera Thysanus and Hadrophyllia are free living, as are all the extinct species of Manicina. See Additional file 5 for stratigraphic references. (*Favia favioides range extends to 65.5 Ma – not shown). origination time of D. labyrinthiformis is pushed back (particularly Thysanus and Manicina) is confirmed by approximately 1.6 MY earlier than previously seen in the both the fossil record and molecular phylogeny. Lastly, fossil record, putting it closer to the origination times of the appearance of new reef dwelling species of Favia and the other species. All mean species origination dates Diploria is simultaneous in the fossil record around occur shortly prior to the final closure of the Central 5 Ma. American Isthmus at 4.25 - 3.45 Ma [11], but we note that the youngest part of 95% HPD for F. fragum and D. labyrinthiformis overlap with this estimated age of final Discussion closure. The timing of the Manicina node is considerably Phylogenetic relationships within modern Caribbean earlier than the appearance of the Manicina areolata in corals the fossil record, indicating that this genus diverged earl- Thorough sampling of individuals within species in our ier than the first appearance of M. areolata. Deeper combined phylogenetic analysis confirms that most nodes in the tree had significantly larger HPD confidence modern Caribbean species form well-supported mono- intervals, due to the lack of fossil calibrations for earlier phyletic lineages (Figure 1). This allows us to reject the taxa. The estimate of origination time for the ingroup idea that widespread hybridization on ecological time was 14.10 Ma (95% HPD interval: 8.77-20.09), while ori- scales [42] is important to the evolution of Caribbean gination of the entire Caribbean Faviidae group is indi- faviids, though limited introgression not detected by this cated by the root node at 17.56 Ma (95% HPD interval: data set might have played a creative role in adaptive 10.04-26.44). These dates coincide with the timing of the processes [43]. The exception lies within the two mod- closure of the Tethys Sea in the eastern Mediterranean ern species of Manicina, M. areolata and M. mayori, [10]. where extensive allele sharing between species might in- Overlay of the molecular phylogeny onto the fossil dicate sub-species status. While M. areolata is a sea- stratigraphy reveals three striking patterns (Figure 2B). grass specialist, and drifts remarkable distances on the First, older and extinct Diploria and Favia cannot be sediment surface as a free-living adult [44], M. mayori is reconciled with this molecular tree, suggesting these gen- a rare reef species that remains permanently attached as era are not monophyletic. Second, the origination and di- an adult. In Panama, these two morphologically distinct versification of a clade of sediment dwelling corals species co-occur within sites yet segregate ecologically by depth-related habitats. An approach that combines Table 1 Divergence dates estimated from BEAST ecological/reproductive comparisons, morphometric data, and further genetic analyses such as the Taxa Date of Origination - Ma coalescent-based model of isolation and migration [45] Mean Median 95% Highest Posterior Density (HPD) interval could resolve this issue. Above the species level, we could not further resolve C. natans 6.25 6.01 5.16-8.02 the branching order of species within the larger clade. F. fragum 5.74 5.53 3.52-8.36 Previous single locus phylogenies using mitochondrial D. clivosa 5.60 5.41 4.67-7.06 and nuclear genes that have included this group have D. labyrinthiformis 4.66 4.37 3.01 - 7.06 shown a similar lack of resolution within the Caribbean D. strigosa 6.03 5.76 4.67-8.08 faviids[35,36]. Another study by Nunes et al. [39] shows Manicina 8.21 7.97 4.81-12.08 some supported structure within this group. However, as this paper was looking mainly at broader scale phylogeo- Ingroup 14.10 13.70 8.77-20.09 graphic relationships, sampling was done on only few Root 17.56 16.86 10.04-26.44 individuals per species within the Caribbean faviids and Schwartz et al. BMC Evolutionary Biology 2012, 12:123 Page 6 of 10 http://www.biomedcentral.com/1471-2148/12/123

using only a single mitochondrial marker and a single Sea (Figure 2A). While these dates support the widely nuclear marker. For examining relationships below the accepted notion of divergence driven by increased isola- familial level, the low rates of mtDNA evolution in corals tion of the region, the radiation of the stem group is might limit the ability to detect more complex topologies much later than indicated by the fossil record amongst these species. With the increased sampling sizes (Figure 2B). The origination of the Favia-Diploria- of multiple loci with higher levels of variation (Add- Manicina (FDM) clade is in the early Miocene, but older itional file 2), we found little evidence for Oligocene Diploria fossils and Eocene Favia fossils are within genera, and branch lengths tended to be long more distantly related, suggesting that both genera are (Figure 1). Therefore, our inability to resolve relation- para- or polyphyletic. Ken Johnson reached a similar ships among species is consistent with rapid diversifi- conclusion based on morphological differences [22], hy- cation and short internal branch lengths deeper in the pothesizing that early Diploria and Favia are unrelated tree. With the rapidly declining cost of high through- to their modern morphological counterparts. Morpho- put sequencing, a phylogenomic approach [46,47] for logical convergence appears to be a common theme in this set of taxa is likely to improve topological coral evolution [35] and our analysis points out some of resolution. the difficulties in determining the systematic positions of extinct taxa. The use of more informative micro- Fossils and molecules reveal the tempo and mode of structural characters that can be quantified in both living Caribbean coral diversification and fossil species may be a promising approach to this Molecular divergence dating indicates extant Caribbean problem [30]. “faviid” corals radiated rapidly during the late Miocene Congruence of morphology, stratigraphy, and esti- to early Pliocene (Figure 2). This ecological radiation mates of node ages can be used to include fossil taxa into coincides with a series of biological and physical changes potentially monophyletic lineages. For example, the di- in the structure of shallow marine habitats during the verse members of living and fossil taxa of the genus early geological development of the Isthmus of Panama. Manicina form a well-supported monophyletic group in During the Late Miocene, shallow marine habitats were Johnson’s morphological phylogeny with all fossil origin- dominated by broader and more gently sloping sedi- ation dates falling within the lower confidence interval mentary shelves [48], while productivity in the water col- for the molecular Manicina node age (Figure 2A). Super- umn above was much higher compared to the modern imposing the age-calibrated molecular phylogeny onto productivity of the Caribbean Sea [49]. Klaus et al. [15] stratigraphy significantly alters the interpretation of the hypothesize that these extensive mesophotic sedimentary speed of evolution in this group (Figure 2B), indicating bottoms may have selected for free-living coral species rapid diversification of sediment dwelling corals in the with large morphologies that were efficient at late Miocene. heterotrophic feeding. Interestingly, our node age for the clade containing the two living Manicina species is Are punctuated patterns driven by adaptation? 8.21 Ma, which coincides with the appearance of other Our time calibrated phylogeny confirms fossil evidence sibling Manicina species in the fossil record that have that extant Caribbean coral species originated during a since gone extinct [22]. Thus, it appears we are sampling period of lineage diversification between 4 and 6 Ma the evolutionary remnants of a once more diverse and (Figure 2). This diversification event corresponds with ecologically dominant clade. As the Miocene transitions ecological radiation into three main ecological niches ex- into the Pliocene, the increasingly isolated Caribbean Sea emplified by modern Caribbean faviids [21]: (i) small, becomes more oligotrophic and the once broad shelf free living morphologies adapted to sedimentary environ- habitats are now dominated by steeper reef platforms, ments (ii) attached species that live in shallow rubble ideal conditions for primarily photoautotrophic reef beds and patch reefs, and (iii) massive colonies the build species. Our time-calibrated phylogeny shows repeated forereef slopes (Figure 1). During the same period, we speciation events of Diploria and Favia species be- also see diversification of reproductive strategies [23], tween ~ 4 – 6 Ma that are either reef specialists or are from tightly synchronized annual mass-spawning events limited to very shallow (< 5 m) seagrass habitats. Thus and broadcasting larvae typical of Diploria [24] to mul- the fossil record and molecular data broadly agree on tiple lunar cycles of reproduction and brooding develop- the timing of these ecological radiations, which are ment found in Favia and Manicina [50,51]. temporally correlated with changes in habitat structure The changes in morphology and life history coupled and productivity. with widespread environmental changes are suggestive Deeper in the tree, node ages for the stem groups of that diversification of Atlantic “faviid” coral might be the Caribbean faviids correspond to the isolation from driven by the evolution of adaptive traits. Using our the Mediterranean during the closure of the Tethys current phylogeny as a stepping stone, increased genomic Schwartz et al. BMC Evolutionary Biology 2012, 12:123 Page 7 of 10 http://www.biomedcentral.com/1471-2148/12/123

and taxonomic sampling of Atlantic corals should allow described in Carlon and Lippé [59]. Skeletal vouchers us to take advantage of several promising new approaches were processed by bleaching in a 50% hypochlorite/water to estimate rates of diversification and evaluate models of solution overnight, rinsing in DI water, and thor- adaptive radiation [52,53]. oughly drying. Species identification was conducted by D. Carlon in the field and confirmed by A. Budd from Conclusions vouchers. Complete descriptions of these taxa, photos, By combining data from the fossil record with molecular and references are available from the Neogene Marine phylogenetic techniques for the first time, this study has Biota of Tropical American (NMITA) database (http:// given us extensive insight into the tempo of diversifica- .geology.uiowa.edu/). tion in an ecologically diverse group of Caribbean corals. Two separate lines of evidence now verify the existence Laboratory protocols of a Mio-Pliocene radiation, while we have been able to For this study we chose to focus on nuclear markers, additionally confirm species identity, verify origination since rates of mitochondrial DNA evolution have been dates, and understand taxonomic relationships in this di- shown to be very slow in corals, limiting the ability to verse and ecologically important group. These findings detect more recent speciation events [60]. We amplified give us the tools to re-interpret trends seen in the fossil three single-copy nuclear regions with primers listed in record, allowing us to begin to link patterns of macro- Additional file 4. Pax-C and CaM primers target introns evolution to paleoenvironmental changes and gain a new located within the Pax protein and calmodulin binding comprehension into the origins and drivers of diversity protein respectively[61,62], while MaSC-1 is an anonym- in the Caribbean. ous region originally sequenced in annu- Besides clarifying evolutionary history, this study has laris [63]. For PCR amplification of all three loci, we broader contemporary implications. With global change combined: 1 μl of 1x to 100x diluted genomic DNA with currently causing a rapid decline in popula- a24μl PCR master mix consisting of: 0.3 μl of each pri- tions around the world [54,55], understanding the pro- mer (10 μM), 1 μl dNTPs (2.5 mM each), 2.5 μl 10x re- cesses that generated diversity in coral species will be action buffer, 1 μlMgCl2 (25 mM), 1 μl BSA (10 mg/ml), key to predicting future changes and directing conserva- 0.3 μl Taq polymerase (Bioline), and 17.6 μlofH20. Each tion efforts [56]. It has been suggested that species that reaction was run at 95 °C for 10 min, 30 cycles of 94 °C evolved in a more heterogeneous environment and sur- for 30s, Ta for 40s and 72 °C for 1 min, with a final ex- vived past climatic fluctuations will be more resistant to tension of 72 °C for 10 minutes. Purified PCR products current global change [29]. Understanding patterns of were sequenced on ABI 3731 XL 96 capillary DNA ana- Caribbean coral evolution during the Pleistocene may be lyzers at the University of Hawaii at Manoa and chroma- key to understanding the potential outcomes of current tograms were then analyzed and edited using environmental impacts. Sequencher 4.5 (Gene Codes). Direct reads revealed indels segregating within many of the species, and pre- Methods liminary verified multiple indels within all three Taxon sampling genes. Since phasing length-variant heterozygotes (LVHs) We sampled six of the seven nominal species from the from direct reads proved unreliable, we cloned 86% of genera Favia, Diploria, and Manicina that form a mono- the PCR products from individuals with LVH pheno- phyletic group within the Caribbean Faviidae [35,39]. types. We cloned PCR products using a TOPO TA clon- The single missing taxon is Favia gravida, closely related ing kit (Invitrogen) and sequenced using standard M13 to Favia fragum, but with a distinct non-Caribbean dis- primers. Single nucleotide polymorphism heterozygotes tribution in that it has been only described from Brazil were phased using the software PHASE v2.1.1 [64,65]. and West Africa [57,58]. We used the genus Colpophyl- We used the non-recombination model, and phase lia as the outgroup because it has previously been shown thresholds of 0.90. To convert data between FASTA and to be a stem taxon to the ingroup species [35,39]. Exten- PHASE formats, we used the webtool SeqPHASE [66]. sive sampling within each species was conducted at two Haplotype sequence data are available as Genbank Pop- reef systems in the Caribbean Sea: the Bocas del Toro, Sets. Accession numbers are listed by species in Add- Panama, and the Florida Keys, USA with additional F. itional file 1. fragum sampled from St. Croix, USVI. The complete list of samples and collection localities is provided in Add- Phylogenetic analyses itional file 3. Skeletal vouchers are deposited in the Uni- Gene trees versity of Iowa Paleontology Repository (http:// Allele sequences were aligned automatically using geoscience.clas.uiowa.edu/paleo/index). Samples were MAFFT v6 [67,68], and corrected by eye in MacCla- collected, preserved, and genomic DNA extracted as dev4.08 [69]. Indels were coded as missing data. Models Schwartz et al. BMC Evolutionary Biology 2012, 12:123 Page 8 of 10 http://www.biomedcentral.com/1471-2148/12/123

for molecular evolution for Bayesian analysis for both Table 2 Stratigraphic ranges, BEAST calibrations, and gene trees and the partitioned tree were selected using section references the Akaike Information Criteria (AIC) in jModelTest Species Fossil Date of Node Calibration Median (95% v0.1.1 [70,71]. For the gene trees, the best fitting model 1st occurrence (mean, standard interval) (Ma) deviation, offset) for the CaM and Pax-C alignment was GTR + G, and for MaSC-1 the model was GTR. Bayesian trees for all three C. natans 5.1-5.3 0.5,1,5.1 6.7 (5.3-16.8) loci were generated in MrBayes v3.1 (5,000,000 genera- F. fragum 3.0-5.6 1.1,0.8,3.0 6.0 (3.8 - 17.4) tions, nruns = 2, nchains = 4) [72,73]. Trees were sampled D. clivosa 4.6-5.9 0.6,1,4.6 6.4 (4.9-17.5) every 100 generations and 5,000 trees were discarded as D. labyrinthiformis 2.9-3.1 0.7,1,2.9 4.9 (3.2-17.2) burn-in. Maximum likelihood analysis was performed D. strigosa 4.6-5.9 0.6,1,4.6 6.4 (4.9-17.5) using RaxML 7.2.6 [74,75] with 1000 rapid bootstraps M. areolata 3.0-5.6 1.1,0.8,3.0* 6.0 (3.8 - 17.4) using the default GTR + G model for all loci at the recommendation of the programmers. M. mayori 2.9-3.1 n/a* Ranges in fossil dates of first occurrence reflect accuracy of section dating. References for dates can be found in Additional file 5. *Calibration for Manicina Partitioned trees is at genus node since species not well supported in phylogeny. Individuals sequenced at all three loci were used for the construction of a combined partitioned tree. For hetero- natans and the genus node for Manicina (Table 2). Cali- zygotes, SNPs were coded as ambiguous data using brations of nodes were done following the guidelines of standard IUPAC nucleotide ambiguity codes. For the Ho and Phillips [79]. For all date priors, we used a log- Bayesian analysis, the K80 + G model was chosen for the normal distribution with a hard minimum bound set at CaM partition, the HKY + I + G model for theMaSC-1 youngest possible date of first appearance in the fossil partitions, and the HKY model for the Pax-C partition. record. The mode of the distribution was set to be Bayesian trees were generated using MrBayes v3.1 slightly older than the oldest possible date of first ap- (20,000,000 generations, nruns = 4, nchains = 4). Trees pearance. Finally, the 95% probability distribution was were sampled every 1000 generations and 5000 trees set to encompass a soft maximum bound at the time of were discarded as burn-in. Maximum likelihood analysis the closure of the Tethys (~17 Ma). These distributions was performed using RaxML v7.2.6 with 1000 rapid incorporate the best-known estimates for origination bootstraps on the Cipres Web Server [76]. For this ana- dates of these taxa, but are wide enough to allow for lysis, the default GTR + G model was used as above. shifts in dates that may reflect errors due to interpret- ation or incompleteness of the fossil record. Divergence dating BEAST was run 4 times (generations = 20,000,000, The program BEAST v1.7.1 [77] was used to estimate diver- samplefreq = 1000) on the Bioportal webserver at the gence dates at species nodes using available fossil data for University of Oslo [80]. Log files were examined in calibration. Input files for the analysis were generated with Tracer v.1.5 [41] to assess convergence of each run. After Beauti v1.5.4 using a partitioned alignment file of 80 indivi- a 10% burn-in was removed, logs and trees for all runs duals. We used a Yule process speciation prior for branch- were combined in LogCombiner v1.7.1 and chronograms ing rates along with an uncorrelated lognormal model for a were generated with Tree Annotater v1.7.1. relaxed molecular clock. Models for molecular evolution for BEAST analysis were selected using the Akaike Information Additional files Criteria (AIC) in MrModelTest v2.2 [78]. The HKY + G model was used for the CaM and MaSC-1 partitions, and Additional file 1: Alleles and Accession Numbers by Species. the HKY + I model for the Pax-C partition. Base frequencies Number of individuals sequenced per species (n), the number of alleles isolated per locus per species, and Genbank accession numbers. All were estimated throughout the analysis. Based on the phylo- species carried unique alleles, except for the two Manicina species. The genetic analysis, all species nodes were constrained to be two last rows give the number of unique alleles in the combined monophyletic except for M. areolata and M. mayori, which Manicina data sets (Manicina spp.) and the combined 6 ingroup species. were constrained at the genus node. Shape parameter priors Additional file 2: Gene trees for (A) CaM, (B) MaSC-1, and (C) Pax-C. Alleles are designated by locus_allele number. Node labels indicate were taken from MrModeltest v2.2 and priors for rates of Bayesian posterior probabilities/Maximum likelihood bootstrap support, -- evolution and Yule birth rates were chosen based on < 50% ML support. Two-tone boxes indicate alleles shared between taxa. defaults narrowed from preliminary runs. All trees produced in MrBayes v3.1 (generations = 5,000,000, nruns = 2, nchains = 4.) The models of evolution were GTR + G for Cam and Pax-C Stratigraphic ranges of extinct and living Caribbean and GTR for MaSC-1. See Additional File 3 for individual genotypes. Faviidae were compiled from the literature and unpub- Additional file 3: Sampling and genotype data for all individual lished sources (Additional file 5). Fossil stratigraphic corals. Samples and multilocus genotypes used in gene and ranges for extant species were used to calibrate species concatenated trees. Heterozygous genotypes that could not be resolved by cloning or PHASE 2.1.1 are indicated as ‘?/?+’; and ‘0’ indicates PCR nodes for Diploria spp., Favia fragum and Colpophyllia Schwartz et al. BMC Evolutionary Biology 2012, 12:123 Page 9 of 10 http://www.biomedcentral.com/1471-2148/12/123

failure or poor sequence quality. The two last columns designate which 11. O'dea A, Jackson JBC, Fortunato H, Smith JT, D'Croz L, Johnson KG, Todd JA: samples were used in the concatenated ML/Bayes trees (Figure 1) and Environmental change preceded Caribbean extinction by 2 million years. – the BEAST analysis (Figure 2). Proc Natl Acad Sci USA 2007, 104(13):5501 5506. 12. Alva-Campbell Y, Floeter SR, Robertson DR, Bellwood DR, Bernardi G: Additional file 4: Primers used for direct sequencing. Sequences and Molecular phylogenetics and evolution of Holacanthus angelfishes annealing temperatures for primers used in this study. (Pomacanthidae). Mol Phylogenet Evol 2010, 56(1):456–461. Additional file 5: Stratigraphic Ranges of the Fossil Caribbean 13. Knowlton N, Weigt LA: New dates and new rates for divergence across Faviidae. Compiled first and last occurrence data, references, and notes the Isthmus of Panama. P Roy Soc Lond B Bio 1998, 265(1412):2257–2263. for all Caribbean fossil faviid species. 14. Lessios HA: The Great American Schism: Divergence of Marine Organisms After the Rise of the Central American Isthmus. Annu Rev Ecol Evol Syst 2008, 39:63–91. Competing interests 15. Klaus JS, Lutz BP, McNeill DF, Budd AF, Johnson KG, Ishman SE: Rise and fall The authors declare they have no competing interests. of Pliocene free-living corals in the Caribbean. Geology 2011, 39(4):375–378. Authors’ contributions 16. Collins LS, Budd AF, Coates AG: Earliest evolution associated with closure SAS designed the study, generated molecular sequence data, ran of the Tropical American Seaway. Proc Natl Acad Sci USA 1996, phylogenetic analyses, and drafted the manuscript. AFB confirmed 93(12):6069–6072. morphological identification of species, compiled stratigraphic data, and 17. Jackson JBC: The future of the oceans past. Philos T R Soc B 2010, helped draft the manuscript. DBC participated in the study design, collected 365(1558):3765–3778. field samples, ran phylogenetic and phasing analyses, and helped draft the 18. O'dea A, Jackson J: Environmental change drove macroevolution in manuscript. All authors read and approved the final manuscript. cupuladriid bryozoans. P Roy Soc B-Biol Sci 2009, 276(1673):3629–3634. 19. Smith JT, Jackson JBC: Ecology of extreme faunal turnover of tropical – Acknowledgements American scallops. Paleobiology 2009, 35(1):77 93. We thank S. Barnes, J. Coloma, J. Fitzpatrick, and S. Walls for help collecting in 20. Todd JA, Jackson JBC, Johnson KG, Fortunato HM, Heitz A, Alvarez M, Jung the field. Brendan Holland kindly provided Diploria samples from the Flower P: The ecology of extinction: molluscan feeding and faunal turnover in – Garden Banks. Collections from St. Croix were funded by a grant from the the Caribbean Neogene. P Roy Soc Lond B Bio 2002, 269(1491):571 577. University Research Expeditions Program, University of California (UREP) and 21. Budd AF, Johnson KG: Origination preceding extinction during late – we thank all expedition participants for excellent field assistance. Allen Chen’s Cenozoic turnover of Caribbean reefs. Paleobiology 1999, 25(2):188 200. lab graciously provided the original Calmodulin primers. Thanks to Joanna 22. Johnson KG: A phylogenetic test of accelerated turnover in Neogene Bince for her assistance in sequencing and cloning introns, and George Caribbean brain corals (Scleractinia: Faviidae). Palaeontology 1998, – Roderick for lab space and computer time for the final portions of the 41:1247 1268. analysis. All corals were collected by permits from the Republic of Panama 23. Baird AH, Guest JR, Willis BL: Systematic and Biogeographical Patterns in (ANAM and CITES) and the Florida Keys National Marine Sanctuary. Our field the Reproductive Biology of Scleractinian Corals. Annu Rev Ecol Evol Syst work in Panama benefited from the extremely helpful staff of the 2009, 40:551–571. Smithsonian Tropical Research Institutes Bocas del Toro Station. This research 24. Weil E, Vargas WL: Comparative aspects of sexual reproduction in the was supported by a grant from the National Science Foundation Caribbean coral genus Diploria (Scleractinia: Faviidae). Mar Biol 2010, (DEB-05-43661) awarded to DBC. 157(2):413–426. 25. Greenstein BJ, Curran HA, Pandolfi JM: Shifting ecological baselines and Author details the demise of cervicornis in the western North Atlantic and 1Department of Environmental Science, Policy & Management, University of Caribbean Province: A Pleistocene perspective. Coral Reefs 1998, California, Berkeley, CA 94720, USA. 2Department of Geoscience, University of 17(3):249–261. Iowa, Iowa City IA52242, USA. 3Department of Biology, University of Hawaii at 26. Mesolella KJ: Zonation of uplifted Pleistocene coral reefs on Barbados, Manoa, Honolulu, HI 96822, USA. West Indies. Science 1967, 156(3775):638–640. 27. Stanley SM: Paleoecology and diagenesis of Key Largo Limestone, Florida. Received: 15 December 2011 Accepted: 25 July 2012 Bull Amer Ass Petrol Geol 1966, 50(9):1927–1947. Published: 25 July 2012 28. Hunter IG, Jones B: Coral associations of the Pleistocene Ironshore Formation, Grand Cayman. Coral Reefs 1996, 15(4):249–267. References 29. Johnson KG, Budd AF, Stemann TA: Extinction Selectivity and Ecology of – 1. Schluter D: The ecology of adaptive radiation. Oxford: Oxford University Press; Neogene Caribbean Reef Corals. Paleobiology 1995, 21(1):52 73. 2000. 30. Budd AF, Romano SL, Smith ND, Barbeitos MS: Rethinking the Phylogeny 2. Gavrilets S: Fitness landscapes and the origin of species. Princeton, NJ: of Scleractinian Corals: A Review of Morphological and Molecular Data. – Princeton University Press; 2004. Integr Comp Biol 2010, 50(3):411 427. 3. Givnish TJ, Systma KJ: Molecular evolution and adaptive radiation. Cambridge 31. Kitahara MV, Cairns SD, Stolarski J, Blair D, Miller DJ: A Comprehensive (United Kingdom): Cambridge University Press; 1997. Phylogenetic Analysis of the Scleractinia (, ) Based on 4. Budd AF, Pandolfi JM: Evolutionary Novelty Is Concentrated at the Edge Mitochondrial CO1 Sequence Data. PLoS One 2010, 5(7):e11490. of Coral Species Distributions. Science 2010, 328(5985):1558–1561. 32. Romano SL, Cairns SD: Molecular phylogenetic hypotheses for 5. Jackson JBC, Cheetham AH: Tempo and mode of speciation in the sea. the evolution of scleractinian corals. Bull Mar Sci 2000, – Trends Ecol Evol 1999, 14(2):72–77. 67(3):1043 1068. 6. Geary DH, Hunt G, Magyar I, Schreiber H: The paradox of gradualism: 33. Romano SL, Palumbi SR: Evolution of scleractinian corals inferred from phyletic evolution in two lineages of lymnocardiid bivalves (Lake molecular systematics. Science 1996, 271(5249):640–642. Pannon, central Europe). Paleobiology 2010, 36(4):592–614. 34. Benzoni F, Stefani F, Stolarski J, Pichon M, Mitta G, Galli P: Debating 7. Hull PM, Norris RD: Evidence for abrupt speciation in a classic case of phylogenetic relationships of the scleractinian : molecular gradual evolution. Proc Natl Acad Sci USA 2009, 106(50):21224–21229. and morphological evidences. Contrib Zool 2007, 76(1):35–54. 8. Hunt G: The relative importance of directional change, random walks, 35. Fukami H, Budd AF, Paulay G, Sole-Cava A, Chen CLA, Iwao K, Knowlton N: and stasis in the evolution of fossil lineages. Proc Natl Acad Sci USA 2007, Conventional obscures deep divergence between Pacific and 104(47):18404–18408. Atlantic corals. Nature 2004, 427(6977):832–835. 9. Gavrilets S, Losos JB: Adaptive Radiation: Contrasting Theory with Data. 36. Fukami H, Chen CA, Budd AF, Collins A, Wallace C, Chuang Y-Y, Chen C, Dai Science (Washington D C) 2009, 323(5915):732–737. C-F, Iwao K, Sheppard C, et al: Mitochondrial and Nuclear Genes Suggest 10. Rögl F: Palaeogeographic considerations for Mediterranean and that Stony Corals Are Monophyletic but Most Families of Stony Corals Paratethys seaways (Oligocene to Miocene). Annalen des Naturhistorischen Are Not (Order Scleractinia, Class Anthozoa, Cnidaria). PLoS One Museums in Wien 1998, 99:279–310. 2008, 3(9):e3222. Schwartz et al. BMC Evolutionary Biology 2012, 12:123 Page 10 of 10 http://www.biomedcentral.com/1471-2148/12/123

37. Huang D, Licuanan WY, Baird AH, Fukami H: Cleaning up the 'Bigmessidae': 61. Catmull J, Hayward DC, McIntyre NE, Reece-Hoyes JS, Mastro R, Callaerts P, Molecular phylogeny of scleractinian corals from Faviidae, , Ball EE, Miller DJ: Pax-6 origins-Implications from the structure of two and Trachyphylliidae. BMC Evol Biol 2011, 11:37. coral Pax genes. Dev Genes Evol 1998, 208(6):352–356. 38. Huang DW, Meier R, Todd PA, Chou LM: More evidence for pervasive 62. Chiou C-Y, Chen IP, Chen C, Wu HJ-L, Wei NV, Wallace CC, Chen CA: paraphyly in scleractinian corals: Systematic study of Southeast Asian Analysis of calmodulin (CaM) indicates that Faviidae (Cnidaria; Scleractinia) based on molecular and morphological scleractinian corals possess the ancestral Exon/Intron organization of the data. Mol Phylogenet Evol 2009, 50(1):102–116. eumetazoan CaM gene. J Mol Evol 2008, 66(4):317–324. 39. Nunes F, Fukami H, Vollmer SV, Norris RD, Knowlton N: Re-evaluation of the 63. Severance EG, Szmant AM, Karl SA: Single-copy gene markers isolated systematics of the endemic corals of Brazil by molecular data. Coral Reefs from the Caribbean coral, Montastraea annularis. Mol Ecol Notes 2004, 2008, 27(2):423–432. 4(2):167–169. 40. Forsman ZH, Barshis DJ, Hunter CL, Toonen RJ: Shape-shifting corals: 64. Stephens M, Smith NJ, Donnelly P: A new statistical method for haplotype Molecular markers show morphology is evolutionarily plastic in . reconstruction from population data. Am J Hum Genet 2001, BMC Evol Biol 2009, 9. 68(4):978–989. 41. Rambaut A, Drummond AJ: Tracer v.1.5.; 2007. Available from http://beast. 65. Stephens M, Scheet P: Accounting for decay of linkage disequilibrium in bio.ed.ac.uk/Tracer. haplotype inference and missing-data imputation. Am J Hum Genet 2005, 42. Veron JEN: Corals in space and time: the biogeography and evolution of the 76(3):449–462. Scleractinia. Ithaca: Comstock/Cornell; 1995. 66. Flot JF: SEQPHASE: a web tool for interconverting phase input/output 43. Arnold ML: Natural hybridization and evolution. New York: Oxford University files and fasta sequence alignments. Mol Ecol Resour 2010, 10(1):162–166. Press; 1997. 67. Katoh K, Misawa K, Kuma K, Miyata T: MAFFT: a novel method for rapid 44. Johnson KG: Population-Dynamics of a Free-Living Coral - Recruitment, multiple sequence alignment based on fast Fourier transform. Nucleic Growth and Survivorship of Manicina-Areolata (Linnaeus) on the Acids Res 2002, 30(14):3059–3066. Caribbean Coast of Panama. J Exp Mar Biol Ecol 1992, 164(2):171–191. 68. Katoh K, Toh H: Recent developments in the MAFFT multiple sequence – 45. Hey J, Nielsen R: Integration within the Felsenstein equation for improved alignment program. Brief Bioinform 2008, 9(4):286 298. Markov chain Monte Carlo methods in population genetics. Proc Natl 69. Maddison DR, Maddison WP: MacClade 4: Analysis of phylogeny and Acad Sci USA 2007, 104(8):2785–2790. character evolution. Version 4.0. Sunderland: Sinauer Associates; 2000. 46. Rokas A, Williams BL, King N, Carroll SB: Genome-scale approaches to 70. Guindon S, Gascuel O: A simple, fast, and accurate algorithm to estimate – resolving incongruence in molecular phylogenies. Nature (London) 2003, large phylogenies by maximum likelihood. Syst Biol 2003, 52(5):696 704. 425(6960):798–804. 71. Posada D: jModelTest: Phylogenetic model averaging. Mol Biol Evol 2008, – 47. Rokas A, Carroll SB: More genes or more taxa? The relative contribution of 25(7):1253 1256. gene number and taxon number to phylogenetic accuracy. Mol Biol Evol 72. Huelsenbeck JP, Ronquist F: MRBAYES: Bayesian inference of phylogenetic – 2005, 22(5):1337–1344. trees. Bioinformatics 2001, 17(8):754 755. 48. McNeill DF, Eberli GP, Lidz BH, Swart PK, Kenter JAM: Chronostratigraphy of 73. Ronquist F, Huelsenbeck JP: MrBayes 3: Bayesian phylogenetic inference – prograding margins: A record of sea-level changes under mixed models. Bioinformatics 2003, 19(12):1572 1574. and dynamic slope sedimentation, western Great Bahama Bank. 74. Stamatakis A: RAxML-VI-HPC: Maximum likelihood-based phylogenetic Subsurface Geology of a Prograding Carbonate Platform Margin, Great analyses with thousands of taxa and mixed models. Bioinformatics 2006, – Bahama Bank: Results of the Bahamas Drilling Project: SEPM, Special 22(21):2688 2690. Publication 2001, 70:101–136. 75. Stamatakis A, Hoover P, Rougemont J: A Rapid Bootstrap Algorithm for the – 49. Jain S, Collins LS: Trends in Caribbean Paleoproductivity related to the RAxML Web Servers. Syst Biol 2008, 57(5):758 771. Neogene closure of the Central American Seaway. Mar Micropaleontol 76. Miller MA, Pfeiffer W, Schwartz T: Creating the CIPRES Science Gateway for 2007, 63(1–2):57–74. Inference of Large Phylogenetic Trees.InProceedings of the Gateway Computing Environments Workshop (GCE): 14 Nov. 2010; New Orleans, LA. 50. Johnson KG: Synchronous planulation of Manicina areolata (Scleractinia) 2010:1–8. http://www.ngbw.org/ee/index.php/portal/cite_us. with lunar periodicity. Mar Ecol Prog Ser 1992, 87(3):265–273. 77. Drummond AJ, Rambaut A: BEAST: Bayesian evolutionary analysis by 51. Szmant-Froelich A, Reutter M, Riggs L: Sexual Reproduction of Favia- sampling trees. BMC Evol Biol 2007, 7:214. Fragum Lunar Patterns of Gametogenesis Embryogenesis and 78. Nylander JAA: MrModeltest v2.: Evolutionary Biology Centre, Uppsala Planulation in Puerto-Rico. Bull Mar Sci 1985, 37(3):880–892. University; 2004. http://www.abc.se/~nylander/mrmodeltest2/mrmodeltest2. 52. Simpson C, Kiessling W, Mewis H, Baron-Szabo RC, Müller J: Evolutionary html. diversification of reef corals: a comparison of the molecular and fossil 79. Ho SYW, Phillips MJ: Accounting for Calibration Uncertainty in records. Evolution 2011, 65(11):3274–3284. Phylogenetic Estimation of Evolutionary Divergence Times. Syst Biol 2009, 53. Yoder JB, Clancey E, Des Roches S, Eastman JM, Gentry L, Godsoe W, Hagey 58(3):367–380. TJ, Jochimsen D, Oswald BP, Robertson J, et al: Ecological opportunity and 80. Kumar S, Skjaeveland A, Orr RJS, Enger P, Ruden T, Mevik BH, Burki F, Botnen the origin of adaptive radiations. J Evolution Biol 2010, 23(8):1581–1596. A, Shalchian-Tabrizi K: AIR: A batch-oriented web program package for 54. Hughes TP, Baird AH, Bellwood DR, Card M, Connolly SR, Folke C, Grosberg construction of supermatrices ready for phylogenomic analyses. BMC R, Hoegh-Guldberg O, Jackson JBC, Kleypas J, et al: Climate change, human Bioinforma 2009, 10. impacts, and the resilience of coral reefs. Science 2003, 301(5635):929–933. 55. Pandolfi JM, Bradbury RH, Sala E, Hughes TP, Bjorndal KA, Cooke RG, doi:10.1186/1471-2148-12-123 McArdle D, McClenachan L, Newman MJH, Paredes G, et al: Global Cite this article as: Schwartz et al.: Molecules and fossils reveal trajectories of the long-term decline of coral reef ecosystems. Science “ ” – punctuated diversification in Caribbean faviid corals. BMC Evolutionary 2003, 301(5635):955 958. Biology 2012 12:123. 56. Van Oppen MJH, Gates RD: Conservation genetics and the resilience of reef-building corals. Mol Ecol 2006, 15(13):3863–3883. 57. Boekschoten GJ, Best MB: Fossil and Recent Shallow Water Corals from the Atlantic Islands Off Western Africa Cancap-Contribution No. 56. Zoologische Mededelingen (Leiden) 1988, 62(5–9):99–112. 58. Castro CB, Pires DO: Brazilian coral reefs: What we already know and what is still missing. Bull Mar Sci 2001, 69(2):357–371. 59. Carlon DB, Lippé C: Estimation of mating systems in Short and Tall ecomorphs of the coral Favia fragum. Mol Ecol 2011, 20(4):812–828. 60. Shearer TL, van Oppen MJH, Romano SL, Woerheide G: Slow mitochondrial DNA sequence evolution in the Anthozoa (Cnidaria). Mol Ecol 2002, 11(12):2475–2487.